FOUR PLASTIC SQUARES and some wire are fashioned into a flat vacuum chamber.

Energetic protons from deep space continuously bombard our planet and strike atoms in the upper layers of the atmosphere. These collisions produce tiny nuclear explosions, which in turn give rise to every species in the particle zoo--protons, neutrons, electrons, muons, lambda particles, you name it. Of these, only the muons have enough penetrating power to reach the ground unscathed. Still, the flux of these subatomic particles, known as cosmic rays, is surprisingly high: about 200 rain down on each square meter of Earth every second.

With the instrument described here, ambitious amateurs can monitor the intensity of cosmic rays throughout the day, chart their distribution in the sky and learn something about their energies. The detector consists of two large, flat Geiger counters linked together with a simple electronic circuit. Here's how they work. A set of fine wires carries about 1,000 volts or so. This potential creates an enormous electric field (more than one million volts per meter) near each wire. When a cosmic ray enters this space, it strips some of the atoms in the surrounding gas of a few electrons, which then move toward the nearest positively charged wire. On the way, these electrons gain enough energy from the huge electric field to knock more electrons from other gas molecules. These charges also accelerate and collide to release still more electrons, and so forth.

Within just millionths of a second, the few electrons originally liberated by the passage of the cosmic ray trigger an electric avalanche, causing more than a billion negative charges to cascade down onto the wire. This current flows into a capacitor (C1 on the diagram on page 87), which in turn generates a voltage pulse that feeds into the counting circuitry.

Most Geiger counters are filled with a noble gas, usually helium or argon. Both can be found at welder-supply shops. Helium, so useful for filling balloons, can also be obtained cheaply at any party-goods dealer. Ordinary air also works, albeit at a higher operating voltage.

No matter what gas you're using, you must reduce the pressure in the chamber to about seven centimeters of mercury--about 10 percent of atmospheric pressure. The March 1960 and October 1996 installments of this column describe homemade vacuum pumps that should serve nicely. But you can also reduce the chamber pressure with a bicycle pump if you modify it appropriately (consult the Web site of the Society for Amateur Scientists for details).

Begin construction by cutting four pieces (as shown above) from a rigid sheet of plastic that is 3/8 of an inch, or about one centimeter, thick. Using the edge of a small file, carve a series of small notches spaced precisely half a centimeter apart on opposite sides of the piece indicated in the diagram. Next, arrange a length of hefty “bus wire” (solid copper wire without insulation) as shown. Secure it with tape at the corners and apply tiny dollops of five-minute epoxy between the notches. Also add a liberal amount of epoxy to the wire along the side you've not filed, making sure to leave at least one centimeter around the perimeter untouched to accommodate the piece that fits above.

Use the notches to position the “sense wire,” bare copper wire that is only 10 thousandths of an inch (about 250 microns) thick. Wrap this fine wire around the square plastic frame, using a steady hand to maintain tension, and hold the ends in place temporarily with duct tape.

Now you must delicately solder the sense wire to the bus wire everywhere they touch. Use a hot soldering iron and plenty of flux. Then attach the sense wire to the frame with a liberal coating of slow-setting (24-hour) epoxy. Once it sets, carefully snip the excess wire just where it emerges from the epoxy to yield a single plane of 29 sense wires. Solder a high-voltage lead to the bus wire. Use epoxy to attach aluminum foil to the top and bottom plastic squares, as shown in the diagram, and install the stopcock and low-pressure gauge to the top piece. Solder ground wires to the aluminum foil. Carve three narrow channels in the middle plastic pieces for the high-voltage and ground wires. (A Dremel tool will work well.)

HIGH-VOLTAGE SOURCE and simple circuitry detect the passing cosmic-ray particles.

Now you're ready to assemble the chamber. The unit has to be airtight, so run continuous beads of silicone aquarium cement where the layers join and put a heavy weight on top or clamp things while the adhesive sets. Make sure also that the channels that hold the lead wires are especially well sealed. The speaker provides an audible output, but for more exacting work I count my cosmic rays on a digital pedometer that I bought for $15 from Radio Shack (catalogue no. 63-618). When this gadget is jostled enough to swing its tiny magnetic pendulum to one side, the magnet pulls together two fine metal strips inside an encapsulated switch, completing a circuit. By bypassing the switch with one of your own, the pedometer can be made to count almost anything that is not producing events too often. The limit seems to be about five times a second, which is just fine for counting cosmic rays. To convert the pedometer, remove the battery cover and nip away at the extreme right side of its plastic case using a small pair of pliers. Then cut off the exposed switch and solder on leads from the detector circuit.

You'll need a variable high-voltage supply to operate the apparatus. Before you power things up for the first time, be absolutely certain that no high-voltage wires are exposed and be extremely careful to avoid any possibility of a dangerous shock. When you're sure that everything is safe, apply 600 volts to start and slowly raise the potential until you just begin to register counts. This setting is your chamber's threshold voltage. The count rate will rise with the applied potential until essentially all the ionizing particles that enter the chamber are detected. At that point (about 1,200 volts for my detector), the count rate levels off. This “plateau” should extend for several hundred volts. As you raise the voltage even higher, secondary effects generate spurious counts, and so the rate rises again. Set your operating voltage at the center of the plateau.

Once you've built and tested two identical chambers, it's easy to construct a cosmic-ray telescope. Just align the two chambers and flip the switch to the A-and-B position, which counts just the events that trigger both detectors. Because particles produced by radioactive decay don't have enough energy to pass through both plastic boxes, your telescope will now show only cosmic rays. This equipment affords many opportunities for research. Position the chambers close together to detect daily and seasonal variations in the flux of cosmic rays. Or place the detectors farther apart to restrict the angular acceptance of the telescope. This maneuver allows you to measure the flux coming from a given direction and to observe how the rate depends on elevation angle and azimuth. By placing material between the two chambers, you can screen out low-energy cosmic rays. Muons lose about two million electron volts (MeV) of energy for each centimeter of water they pass through. A brick, which is about two times as dense as water, will extract about 4 MeV for each centimeter of thickness. You can use this effect to investigate the energy spectrum of the more feeble muons impinging on your detector. And you can detect the immense “air showers” that very energetic protons spawn by comparing results from two telescopes situated about 100 yards (or meters) apart. With a little imagination and effort, you will surely make some fascinating discoveries.

The Society for Amateur Scientists will offer a kit for this project until January 2002. The package contains only the various electronic components required (apart from the pedometer) and a spool of fine sense wire. The cost is $30. To order, call the society at 401-823-7800. For an ongoing discussion about this project, surf over to www.sas.org and click on the Forum button. You can write the society at 5600 Post Road, #114-341, East Greenwich, RI 02818. To purchase Scientific American's CD-ROM containing every article published in this department through the end of 1999 (more than 1,000 projects in all), consult www.tinkersguild.com or dial toll-free: 888-875-4255. Erratum: Mercury's freezing temperature was incorrectly given in the Amateur Scientist for December 2000. The correct value is -38.9 degrees; the corresponding output voltage in the table on page 104 should read -0.365 volt.